• 1

    posted a message on New to hearthstone, general doubt regarding crafting

    also have you ever wondered why the "most oppressive" decks in hearthstone only have a 55% winrate? it just happened that every single deck in the history of the game is near 50%? given the wide variety of player skill, deck formulas, and all the other factors, it truly is mindboggling that 50% is somehow the magic number. its almost like skill doesnt matter, 50% does... :)

    I just don't think that's that unlikely an outcome without the idea that there's a hypothetical "deck rating" being applied to all levels of matchmaking.  There are a bunch of effects that tend to balance things out without that:

    • Blizzard's design has leaned toward multiple viable archetypes which tend to be either extremely strong or extremely weak against each other.  Thus, if one archetype seems to be dominating (Jade Druid in the first few days of this expansion) a large number of players will gravitate to its obvious counters REALLY FAST, countering its dominance.
    • Their long development cycle probably allows Team 5 to catch really broken mechanics and card combinations (the ones that would make one archetype truly dominant) before they go out the door in most cases.
    • While Ranked rating doesn't resemble an MMR system, a matchmaking system that prioritizes same-rank matches above all else will tend to bias results heavily toward 50%.  Those who perform better than the current rank (on average) rapidly move up and out, and get matched with a new population, while those who underperform stay where they are or drop.  Only a 50% win rate keeps people at the same rank, so ranks will quickly be stratified by performance to get the win rate as close to 50% as possible (though with an overall upward migration driven by rank floors canceling out losses and bonus stars.)

    The hsreplay.net people probably could come up with a way to test what you're proposing.  It would be great if they would.  I'm not saying I know for certain you're wrong, I'm just saying that "change your deck and see what happens over the next 20 games" is a REALLY bad test unless it's such an overwhelming effect that adding blah blah card means you only see hunters, ever, every single time.  And, I doubt they're doing something that coarse and unhelpful.

    the worst thing he could possibly do is pull out his wallet and buy dust wads until he can make a jade druid, etc...

    I agree in general (although I do think that there's some value in spending dust to build one relatively strong deck of some kind just to have the good feedback of some ladder success, preferably an archetype that's shown some long-term viability and isn't dependent on specific cards that are going to age out at the next Standard update.)

    Posted in: Standard Format
  • 1

    posted a message on New to hearthstone, general doubt regarding crafting
    Quote from Matrix801 >>
    Quote from Lysenko >>

    There's no good way to build a system that does what you describe.  You're imagining things.

    https://en.wikipedia.org/wiki/Confirmation_bias

     No, he is right, he may be exaggerating, but there is more to the matchmaker than current rank and stars.  If you don't have the formula then picking winners and losers will average out over the population and appear random, but there are undoubtedly 'factors' involved in the match engine.
    Here's a Developer Q&A from January of this year, describing how the ranked matchmaking experience works.  Listen to the entire discussion, because without it you won't have a clear idea for how candid they're being about the details.  The key point is that they describe improving one's deck having a strong impact on ranking, which could not be a side effect of ranked matchmaking if they were artificially fiddling with the matches based on one's deck contents.
    They do state that Casual mode adds tweaks on top of MMR to try to improve the experience for new players, and they do so in enough detail that if Ranked were looking at the contents of your deck, there's no way they wouldn't have at least alluded to it.
    There's a discussion around 10:50 that talks about possible tweaks to the Ranked mode to make the experience better, where they focus entirely on changes to the ranking system and do not mention tweaks to the matchmaking (as they do in the Casual mode discussion.)
    It is possible that since this video was released they've started experimenting with making changes they don't discuss here.  But, EVEN IF they were doing that actively (and I'm sure they are looking at or experimenting with changes of some kind) the idea that you could take a string of 20 or 50 games and infer that a single-card change you made has some effect on matchmaking is ridiculous.  There's just no way to distinguish between that and random variation.  Even with a collection of hundreds of thousands of games and complete information about all the deck lists used across all of them, identifying non-random effects conclusively would be a really hard statistical problem to solve.
    Also, as for Blizzard looking at deck contents for matchmaking, the way a skill-based matchmaking system like this works involves making hundreds or thousands of player/player comparisons in a very short period of time to match players for a particular game.  Using deck contents in just about any way presents all kinds of implementation problems that make it tricky.  Is the deck assessment happening on the client or the matchmaking server?  How does it get transmitted from one to the other after pressing the matchmaking button?  If it's all client side and they just pass along some kind of scalar "deck rating" to the server for performance's sake, how do they collect metrics to know it's working correctly?  These are all problems their developers would need to solve to do this, and solutions wouldn't fit cleanly into Blizzard's existing matchmaking methodology (which, by the way, is almost certainly shared between their games and operated as a central service, making it harder to customize in this way for Hearthstone.)
    Edit:  TL;DR is that they probably weren't doing anything involving looking at your deck list during matchmaking as recently as January.  It's possible they're doing so now, but even if they were, a string of 20 games would be way too few to see it.  It's also a little suspicious that the specific classes Stubblestubble was saying he never saw just happen to be the least-played classes on the Ranked ladder anyway.
     Edit 2:  Also, I realize we may disagree as far as what we believe about this, but thanks, Stubblestubble, for keeping the conversation civil.
    Posted in: Standard Format
  • 0

    posted a message on New to hearthstone, general doubt regarding crafting
    Quote from Stubblestubble >>

    ^ sure, dude, its not like i track my data or anything...

    Distinguishing between random variation and an actual effect would require tens or hundreds of thousands of games played and knowledge about the Hearthstone population's deck choices that nobody but Blizzard has.  The fact that you named the three least-played classes by Vicious Syndicate's stats as the ones you don't see very often is a good sign that you're mistaking trends in what classes people are playing as the matchmaking system manipulating you.
    Posted in: Standard Format
  • 0

    posted a message on New to hearthstone, general doubt regarding crafting

    There's no good way to build a system that does what you describe.  You're imagining things.

    https://en.wikipedia.org/wiki/Confirmation_bias

    Posted in: Standard Format
  • 0

    posted a message on New to hearthstone, general doubt regarding crafting
    Quote from Stubblestubble >>
    ^ I made the deck that specific way to work around the matchmaking system. its extremely rare that it will que into hunters shamans or rogues. 
     That's not at all how the matchmaking system works.
    Edit:  You're not queueing into hunters, shaman, or rogues because nobody's playing them right now, not because of what you put in your deck.  Across all ranks, those are the three lowest-played classes, responsible for only a little under 20% of the players out there, and more like 15% above rank 5.
    Posted in: Standard Format
  • 0

    posted a message on Having a hard time reaching legend :(
    Quote from Maukiepaukie >>

    You don't need to be skilled to get legend. You just need to be willing to invest some time

    Everyone has their own definition of what it means to "be skilled."  Most Hearthstone players don't play at a level that will get them to Legend even with viable decks and unlimited time.
    That said, some decks are a lot easier than others, some decks are a lot faster at grinding out the games, some decks counter the current meta better at any given time, and if you get enough of these to line up and do put in the time, your chances go way up, even if your skill level is marginal.
    Posted in: Standard Format
  • 0

    posted a message on Having a hard time reaching legend :(

    I wound up back at the bottom of rank 5 at least twice in the season I made Legend.  Just focus on learning your deck better and improving your play, and play as much as you can.

    Posted in: Standard Format
  • 0

    posted a message on New to hearthstone, general doubt regarding crafting
    Quote from Firestorm30 >>

    i stopped buying them and began getting gadgetzan and old gods packs. I do own the karazhan adventure(i do invest a little bit now and then), but i have no idea how and what to invest in to make it worth. 

     If you complete the Karazhan adventure,  you will have all the cards from it.  I'd just hang on to them at least until they rotate out of standard.
    As for buying new packs, I would focus exclusively on Un'Goro and Knights of the Frozen Throne if you have a good collection of Classic cards.  Gadgetzan and Old Gods will be rotating out of Standard around the Spring, while Un'Goro and Knights of the Frozen Throne will be around longer.
    Posted in: Standard Format
  • 1

    posted a message on New to hearthstone, general doubt regarding crafting

    I've spent a lot of money on Hearthstone on NA, but for a while I was playing entirely f2p on Europe just to see how quickly I could get myself to a competitive level.

    If you're just starting out, your first goal really should be to have a playable basic/classic deck of every class.  Trump's basic decks are pretty reasonable just as a starting point, and as you get more cards you can swap out for them.  His basic decks are here:  https://sites.google.com/site/trumpdecks/basicdecks

    You should first strive to get all your classes up to level 10 so you have all the basic (free) cards.  This will also earn you some gold and a bunch of packs.  I would then put as much gold as possible into classic packs for a while, since those cards are (mostly) always going to be in Standard.

    Your main goal should be to complete your quests every day to maximize your gold accumulation.  (You'll find this easiest in Casual mode, because you'll very shortly end up facing other people with basic decks due to how matchmaking works in that mode.)  At least at first, only auto-dust your excess duplicates, don't dust cards individually.  Just keep accumulating gold and dumping it into classic packs and trying to improve your decks for each class wherever possible.

    Once you start accumulating dust from duplicates only, maybe a couple thousand, start looking at the inexpensive competitive decks that EveningWood mentions above.  Your goal should be to have a single fairly competitive deck of your choice, and in addition a playable basic-level deck for each class so you can keep pressing ahead on quests.

    When you are putting together your competitive deck, note that some cards you'll want to craft will require completing at least some of the solo adventure content.  You'll need to save up gold to unlock this, or get out your credit card.  I think only One Night in Karazhan is still in Standard.  The advantage of spending gold on this content is that it has a 100% chance of awarding you a full set of some very good cards that you cannot get any other way.

    Good luck!  A few seasons ago, when midrange hunter was doing just a bit better, I had a legend-viable competitive deck put together on my f2p account within a few weeks.  Midrange hunter isn't quite as good now but it still does pretty well up through rank 15, and if you hit that every season, plus do your dailies, you'll be in very good shape to grow your collection.

    Posted in: Standard Format
  • 1

    posted a message on Why Hasn't Blizzard Buffed any Cards?

    Another reason (and after some thought I think this reason is functionally equivalent to Kibler's) is that if you nerf a card, you make all the other cards that could take its place in a deck more valuable, but if you buff a card, you put the focus on that one card, to the exclusion of others, even out of proportion to the buff.  In other words, nerfing a card provides many options but buffing a card puts undue focus on that card, archetypes that benefit from it, and responses to it.  I don't believe the designers want to steer the meta toward single solutions like that.

    Nerfing a card requires making a decision that a particular deck or play style has spun out of control, which can happen.  Buffing a card requires making a decision that what the game needs right now are not all the other options out there, but an emphasis on this one particular way to play (which many players may not have the cards for or be interested in playing.)  It's just hard to see why the game designers would think to themselves "ok what Hearthstone really needs right now is a lot more people playing quest paladin decks."

    Even in a case like Warlock's current state, a single card buff is unlikely to do the job, and if it did, it would leave the class in a very narrow and fragile state.  A better solution is to drop a bunch of interesting cards on the class in a future expansion and see how it shakes out.

    Posted in: Card Discussion
  • 1

    posted a message on Can everyone reach Legend?
    Quote from Excellent20 >>

    I've played hundreds of games and still haven't gotten the dust to rock with most of the top tier meta. "I don't expect players to take more than 300 games" is a very surprising statement. Can you explain further?

     That statement isn't referring to how long it will take to collect the necessary cards.  It's saying that if you have a tier 1 deck right now, you can achieve Legend in about 300 games in a calendar month if you have the ability to achieve a win rate meaningfully above 50% with it in ranks 5-1.
    Posted in: General Discussion
  • 0

    posted a message on Can everyone reach Legend?

    @mayatola:

    Totally agree.  It takes a lot of time spent playing to get to legend even if one is fully consistent and has the 60%+ win rate of a pro.  Playing more marginally just makes it that much more time-consuming, and the variance along the way (with accompanying rises and falls in rank) are enough to discourage many players who might otherwise be able to do so.

    Posted in: General Discussion
  • 0

    posted a message on Can everyone reach Legend?
    Quote from On3_M4n_Army >>

    I'm having a super hard time reaching legend this season. Went from rank 2 to rank 5 like two times already. Currently rank 3.

    Based on some simulations I did (and posted about here) it's possible to have wildly different experiences from month to month even assuming that you have a fixed, relatively high likelihood to win.  If you have a solid but not spectacular 54% win likelihood* (which is above average for all decks in the meta), you'll find you just don't make it some months unless you can consistently get in 1000 games or more.
    *Note that this is not actual win rate, but the underlying chance that you will win each game before it happens.
    Posted in: General Discussion
  • 0

    posted a message on Can everyone reach Legend?
    Quote from DiamondDM13 >>
    Quote from 1xbenx1 >>
     This is a fallacious argument. A bot can play infinite games. A human can not.
     A Bot cannot play infinite games, there is a limited amount of time the bot can use to play, meaning a finite number of games it can play.
     Currently, the fastest decks resolve their games in about 6 minutes.  In a 31-day month, that's 7440 games, not counting downtime.  So, a bot would have to achieve above about a 46% win rate or better to do so in half the 31-day months attempted.
    Posted in: General Discussion
  • 0

    posted a message on Negativity in Hearthstone (Brian Kibler)
    Quote from GrandPatzer >>
     Or the one with the more luck.
     Good thing there's a test for that!  It's called the P-value test...
    Oh wait, we've been over this already. :)
     
    Posted in: Streams and Videos
  • To post a comment, please login or register a new account.